586 



SCIENCE. 



[N. S. Vol. XIV. No. 355. 



The carbon studied was that of the fila- 

 ment of the ordinary incandescent lamp 

 with treated surface, on the one hand, and 

 a similar filament with smoked surface on 

 the other. Temperatures were computed 

 by means of the empirical relation between 

 them and the ratio of the resistances of 

 the hot and cold filament, established by 

 Chatelier. * 



Isochromatic curves showing the rise in 

 intensity with the temperature were plotted 

 for various wave lengths of the visible 

 spectrum. These, taken separately, ex- 

 hibit the form characteristic of bodies pre- 

 viously investigated, such as platinum and 

 the artificially produced * black body.' 



The temperature or its reciprocal as de- 

 manded by the received equations for radi- 

 ation is nearly proportional to the loga- 

 rithm of the intensity. 



The corresponding isothermal curves, 

 however, show the existence of the remark- 

 able peculiarity noted by Nichols f in a 

 recent paper. The selective radiation in the 

 yellow, described by that author, is even 

 more strikingly developed at the higher 

 temperatures covered in these experiments. 

 It exists alike in the case of the' gray-sur- 

 faced carbon and in lamp black and would 

 lead us to classify carbon rather with the 

 metallic oxides than with the black bodies, 

 as regards the laws of radiation. 



2. ' The Distribution of Energy in the 

 Spectrum of the Acetylene Flame ' : George 

 W. Stewart. 



The great value of acetylene not only for 

 illuminating purposes, but also in experi- 

 mental research, makes the study of the 

 distribution of energy in its spectrum of 

 considerable importance. In the work to 

 be described, a mirror spectrometer and 

 fluorite prism were used to produce the 

 spectrum, and the radiometer of Nichols to 

 measure the radiant energy. The mirror- 



* Chatelier, Journal de Physique (3), I., p. 203. 

 t Nichols, PJiysical Bevieiv, Vol. XIII., 1901. 



prism device of Wadsworth was utilized, 

 the main advantage gained being that the 

 radiometer could be kept stationary. The 

 spectrum fell upon a slit mounted directly 

 in front of the fluorite window of the radiom- 

 eter, thus avoiding the use of any lens 

 whatever. 



The source of light used in the work 

 upon the visible portion of the spectrum 

 was a cylindrical acetylene flame from a 

 single-tip burner. This type was used be- 

 cause the intensity per unit area was 

 greater than in the flat flame. The results, 

 plotted in the form of a curve, afibrd defi- 

 nite data concerning the distribution of 

 energy in the visible spectrum of this flame. 



Owing to the difference in the dispersion 

 of the prism, a slight change in the relative 

 positions of the spectrometer and radiom- 

 eter was of greater consequence in the infra- 

 red than in the visible portion of the spec- 

 trum. 



The most striking characteristic of the 

 curve of energy distribution is the set of 

 elevations which are due to the emission 

 bands of the gases of the flame. If the 

 acetylene is pure, the gases to be expected 

 are COg and H^O. According to Paschen, 

 the maximum points of the emission bands 

 of these gases are as follows : 



H2O Spectrum, 

 , 1.46,u, 1.90/i, 2.83fJ.. 



CO2 Spectrum, 



2.71^^ (2.68fi when COj is not dry) and 4.40^. 



The energy curve shows elevations whose 

 maxima agree very closely with these 

 values. Observations of wave lengths 

 longer than S/j. are useless, the error due to 

 stray radiation being so great. 



Similar measurements were made upon 

 the flame of a bunsen burner adapted for 

 the combustion of acetylene. 



In this curve all five emission bands ap- 

 pear and the values of the wave length at 

 the maximum points agree with the values 



